Cracks in PRC girders are important index related to girder stiffness. However they are often closed except loading timing due to the introduced prestress, thus, it is difficult to detect visually. In the vibrationbased evaluation, it is necessary to separate the exciation characteristics and then evaluate the rapid fluctuation of the bridge frequency depending on the amplitude. In this study, the Bayesian TV-ARX method, which is a time-varying system identification method considering external force characteristics, are applied to a measured displacement response of high-speed railway PRC girder when a train passage. It was founded that the instantaneous decrease of bridge frequency that cannot be estimated by the continuous Wavelet transform occurred in the target bridge. In addition, crack opening at the time of the lower amplitude is a main cause and it can reduce the girder bending stiffness by a maximum of approximately 30% during train passage.

This paper presents a 2-D elastodynamic time-reversal method using the topological sensitivity for the defect detection indicator in order to determine the number and position of defects in an elastic solid. The time-reversal method is the method to estimate the defect position and elsewhere from the convergence positions of the time-reversal waves when the received scattered waves by defects are time-reversed, and sent back again into an elastic solid. In general, it is difficult to quantitatively evaluate the convergence positions of the time-reversal waves. To overcome this problem, several researches have been done by us to determine the convergence positions of the time-reversal waves for scalar wave problems. However, these achievements for scalar wave problems cannot be applicable to elastodynamic problems because the waves in a solid show the property of the elastic wave. Therefore, in this research, we extend the previous researches for scalar waves to 2-D elastodynamics, and try to determine the number and positions of defects in a solid by using an elastodynamic time-reversal waves and the topological sensitivity. The ultrasonic non-destructive testing (UT) using a linear array transducer is considered, and the convolution quadrature time-domain boundary element method (CQBEM) for 2-D elastodynamics is utilized to obtain scattered wave ?elds by defects in a solid. As numerical examples, some defect detection results are demonstrated to investigate the performance of the present method.

To increase the resilience of facilities over a wide area, it is important to immediately understand the situation in a wide area at the time of a disaster. This study proposes a model to immediately estimate an entire target facility’s behavior from information obtained from limited observation points based on spatio-temporal data obtained by numerical analysis. Firstly, we proposed a method to identify the optimal placements of observation points to effectively make the model work. Here, the QR decomposition with column pivoting was used to optimize the observation points. Secondly, an example is applied to the settlement process of embankment on liquefied, and it demonstrated the effectiveness of the proposed method by comparing the estimation error from random observation points. We confirm that the optimized observation points reflect the physical characteristics of the numerical analysis through the above consideration.

Micro X-ray CT imaging technology is effectively used in the field of geotechnical engineering as one of the useful tools to visualize the microscopic morphology of soil. On the other hand, the micromechanics model is one of the methods to construct the constitutive law of macroscopic mechanical behavior from the information on the micromorphology of soil. This study aims to obtain information on a wide range of soil micromorphology using super-resolution to derive a practical micromechanical model that considers the statistical nature of soil micromorphology. Here, we compare the three methods through simple numerical experiments and discuss the effectiveness in terms of their accuracy and possible extrapolation.

The objective of this study is to investigate the effect of upscaling on spatial distribution characteristic of hydraulic conductivity and solute dispersion for a stratified porous model. A geostatistical analysis allowed quantification of spatial characteristic for the upscaled models. As a result, the smoothing of spatial distribution due to upscaling reduced the heterogeneity, whereas the spatial correlation scale was preserved. In order to identify the relationship between the degrees of solute dispersion and upscaling, a combined experimental and numerical investigation is conducted. The data from the experiment are utilized to validate a numerical model that is then used to estimate the strength of the dispersion processes (i.e., longitudinal and transverse macrodispersivities) in the upscaled models. It is shown that upscaling does not affect the transverse macrodispersivity, while the longitudinal macrodispersivity is underestimated due to the decrease of heterogeneity during the upscaling process. Moreover, the result suggested that the mass transfer between layers be altered by increase in the layer thickness due to upscaling, leading to the plume spreading in the longitudinal direction.

The parametric instability of wheels moving with constant velocity on a periodically supported rail is in-vestigated. The railway track is modeled by an infinite Timoshenko beam and viscoelastic supports located with constant spacing. The wheelsets of a moving bogie are represented by two point masses. An analytic solution is derived for the quasi-steady-state wheels-track dynamic interaction problem, and homogeneous matrix equations are obtained for the Fourier coefficients of the wheel/rail contact forces. The influences of the ratio of wheelbase to support spacing and of the damping in the rail pad on the parametric instability are discussed.

Many sampling or monitoring plans have been empirically designed. There are some studies on optimal placement that focus on the uncertainty, but few studies consider the impact of exceeding the limit state. The authors have been researching an optimal sampling planning method based on the idea of Value of Information. It is possible to consider the optimal sampling placement that takes into account not only the uncertainty but also the risk of exceeding the limit state. In this research, the soil contamination sampling problem is used as an example to demonstrate the sampling placement by the proposed method, and the two dimensional contaminated area is detected using Kriging, which is a spatial distribution estimation method. We compared the estimations based on the placement by the proposed method and human judgement, and showed that the result by the proposed mentod was better than those by human judgments.

This study is intended to disucuss and compare accuracy of road profile identification methods utilizing measured accelerations on a moving vehicle. The accuracy of road profile identification is discussed focusing on ways of deriving a state-space equation and the parameters to be identified by the inverse problem. Vehicle accelerations obtained both from simulation and field experiment are utilized to the roadway rough-ness profile identification. The identified results are evaluated by an error function and root mean square error (RMSE) of power spectral denstiy (PSD) of identified roadway roughness profiles. It was confirmed that the method which identifies the roadway roughness profile directly by the least square minimization has higher accuracy than that identified by means of the moving force identificxation (MFI) method. In selection of the optimal regularization parameter, the point of shortest distance from the origin of the L-curve provided the optimal one for the direct roadway roughness profile identification method.

Many research reports on bonded repair and reinforcement using carbon fiber reinforced polymer (CFRP) for steel structures deteriorated by corrosion are published. So far, it was already clarified that the that the stress assuming the composite theory of the thickness-reduced plate and CFRP does not agree with the finite element (FE) analysis and the shear lag theory under the axial force. However, the shear lag theory has not been investigated for CFRP bonded repaired steel plates subjected to bending moment. In this study, shear lag theory and FE analysis were carried out for the CFRP bonded repair of the thickness-reduced plate under pure bending, and the mechanical behaviors were clarified. As the results, it was shown that the sectional forces calculated by the composite theory of thickness-reduced plate and CFRP do not correspond with that given by the FE results and shear lag theory depending on the dimensions of defect and thickness of CFRP under the pure bending condition, even if the bonded length of CFRP was sufficient.

This study investigates the propagation characteristics of surface wave traveling in a random heterogeneous medium. For this purpose, ultrasonic measurements are performed on a coarse-grained granite block as a typical randomly heterogeneous medium. In the ultrasonic testing, a line-focus transducer is used to excite ultrasonic waves, whereas a laser Doppler vibrometer is used to pickup the ultrasonic motion on the surface of the granite block. The measured waveforms are analysed in the frequency domain to evaluate the travel-time for each measurement point based on the Fermat’s principle. From the ensemble of travel-times obtained thus, the probability distribution of the travel-time is established as a function of travel-distance.The uncertainty of the travel-time and its spatial evolution are then investigated using the standard deviation of the travel-time as a measure of the uncertainty. As a result, it was found that the uncertainty is approximately proportional to the mean travel-time divided by the square root travel-distance.This is a finding that would be useful for stochastic modeling of the waves in random heterogeneous media.

A computational method was proposed for the free-surface flows intruding between multiple absorbent particles on the basis of the finite volume method. In the proposed method, the solid area of the particle, which is in contact with water, expands due to the swelling of the absorbent particle. The influences of particle swelling and water absorption on flows are considered with velocity and pressure boundary conditions on the particle surfaces. The proposed method was first applied to the lock-exchange problem in porous media, which consist of multiple non-absorbent particles. Second, the swelling of a single absorbent particle was calculated with the present method in saturated and unsaturated conditions. The basic applicability of the proposed method was confirmed through the discussions about the front position of the heavy current in the first calculations as well as the volume conservation of water in the second investigations. In addition,the proposed method was applied to the dam-break flows in the porous media, which consist of multiple absorbent particles. As a result of the numerical experiments, it was demonstrated that the flow patterns intruding the porous media are largely affected by the swelling particles. In particular, it was found there is a possibility that the total amount of water absorption decreases due to the blocking of the flows by the swelling particles even in case the water absorption speed of the particles is sufficiently high.

The distinct element method (DEM) is used for analyzing interaction of colliding particles in various fields. The conventional DEM has an upper limit of time step mainly because of linear oscillatory dynamics if it has high spring stiffness. Therefore it requires greater computational resources. In addition, it has difficulties in deciding the parameters of particles simulating collision of hard materials. In this study we propose an impulse-based DEM with variable temporal resolution for analyzing industrial machines with dynamic hard spheres. This model has less parameters of particles and larger upper limit of the time step than the conventional model. We verify qualitative accuracy of our proposed model by observing behavior of particles in the ball mill, and compare results obtained by the proposed model with experimental and results of conventional model.

Computations of the two dimensional rectangular container filling are conducted by injecting the highly viscous Newtonian fluids, vertically downward through an aperture on the top boundary. In order to simultaneously calculate the low-viscosity gas surrounding the highly viscous liquid, phase-averaged governing equations are used for two immiscible incompressible fluids with different densities and viscosities. The governing equations are discretized with the finite volume method in the collocated grid system. In particular, the estimation methods of viscosity, density and velocities on computational cell boundaries in the discretized governing equations are investigated and their effectiveness was confirmed through the application to the dam-break flows. As a result of the computations of the container filling with the highly viscous liquid, it was confirmed that the proposed method enables us to reasonably predict flow patterns of the highly viscous liquid, such as steady filling, spreading, splashing and buckling, which arise in the specific relationships between Reynolds number and the ratio H/d, where H is the height of the container and d is the inlet width, as reported in the preceding studies.

Generally, it is considered that fatigue life can be divided into two separate contributions: the crack initiation life and crack propagation life. Recent years, the authors studied on a method which predicts the fatigue crack initiation life by using local strain approach and extended to the assessment method of crack propagation life handled as continuous behavior of crack initiation life. The aim of this paper is to introduce a new method for fatigue crack initiation and propagation life prediction of a non-load carrying fillet joint using an unconventional elasto-plasticity model called the Fatigue SS Model (hereafter, FSS model). The FSS model is based on the Subloading Surface theory, which was enriched by including the elastic boundary and cyclic damage concepts for the description of strain softening behavior within macroscopically elastic stress state. The FSS model was used to investigate the inelastic response of the material under different cyclic loading conditions. In this study, finite element analysis was conducted to assess the fatigue crack initiation life and subsequently crack propagation life by extending the assessment method for fatigue crack initiation life of a non-load carrying welded fillet joint by considering a cyclic plasticity accumulation during fatigue loading.

In recent years, many stainless steel (SUS) panel tanks have been installed as for water supply facilities in order to facilitate maintenance and shorten construction time. Although these tanks are designed and constructed according to the latest design standards, there have been so many cases of damage in SUS tanks due to severe earthquakes such as the Tohoku-Pacific Ocean Earthquake and the Kumamoto Earthquake. In this paper, the time history response analysis of a 3000 × 3000 × 3000mm SUS tank is carried out to show local stress behavior which can cause damages. As a result, it is necessary to study more on the relationship between SUS body stiffness in order to avoid damages under earthquake which can modify future design condition.

Monte Calro simulations are carried out to numerically investigate the heterogeneity of concrete in meso-scale. The concrete is regarded as three-phase composite materials composed of mortar, coarse-aggregates, and the interface. A damage model that allows a kind of mesh-free nite element analysis for heterogeneous materials is applied for modeling cohesive crack growth. After presenting the formulation of the damage model, the numerical results are compared with experimental results to verify the applicability of the damage model to the simulation of cohesive crack growth in concrete. Monte Calro simulations are then performed to investigate the inuence of the heterogeneity with respect to the grain size of coarse aggregates in concrete's meso-scale on the fracture properties.

In this study, the H -matrix method is applied to a harmonic balance-boundary element method (HB-BEM). The HB-BEM is a numerical method which deals with steady-state wave scattering by a crack with contact acoustic nonlinearity. The HB-BEM requires high computational costs because it is necessary to simultaneously solve boundary integral equations for several frequencies which are coupled by nonlinear boundary conditions. Since the Newton method is used in the process of solving the nonlinear equations,the computational time for inversion of the Jacobi matrix is predominant. The inversion process is accelerated by means of the H -matrix method, which allows us to use LU decomposition because the Jacobi matrix is ill-conditioned. Through numerical results, the in fluence due to the arrangement of matrix on the acceleration is examined. The computational costs versus the number of unknown variables are also investigated.

In structural analysis, depending on the shape of the objective structure and members, the use of both solid elements and beam elements can improve the computational efficiency. However, it is known that the accuracy of stress distribution near connection is deteriorated for the members such as thin-walled crosssections and composite cross-sections, where cross-sectional deformation cannot be ignored. The reason for the problem is the inconsistency between the displacement fields of the beam element which assumes plane sections remain plane and the solid element. To solve this problem, we propose a method to connect a beam element which allows for the cross-sectional deformation due to shear lag and transverse shear and the solid element. Feasibility of the proposed method is confirmed through a couple of numerical examples by comparing the solution of the method with that of the model with the only solid element.

In this study, a novel implicit stress-update algorithm of the extended subloading surface model, a class of unconventional plasticity model, incorporating the nonhardening strain region is developed to improve the predictive capability for cyclic loading property of metallic materials. Ohno (1982; J. Appl. Mech., ASME) proposed a sophisticated plasticity model equipped with nonhardening strain region in order to reproduce the stagnation of isotropic hardening resulting in transient stabilization of hysteresis loop during cyclic loading with a fixed strain amplitude. In the plastic strain space, when the plastic strain moves inside the nonhardening strain region, the region does not expand, and hence, isotropic hardening does not proceed,while only kinematic hardening proceeds. When the plastic strain moves outside of the nonhardening strain region, isotropic hardening proceeds being accompanied by expansion of the region. As a specific prototype model, von Mises plasticity model is reformulated based on the hyperelastic constitutive equations within the framework of infinitesimal deformation theory. A numerical algorithm for stress calculation using return mapping is developed, in which we devise a novel scheme consisting of hardening/nonhardening judgment and nested iteration to update variables associated with the nonhardening strain region. Numerical examples of cyclic loading demonstrate that the developed constitutive model properly reproduces the stabilization of hysteresis loop. The iteration solution process exhibits satisfactory convergent property, which validates the numerical scheme developed in this study.

In the 2011 earthquake off the Pacific coast of Tohoku, tsunami-induced overflow had resulted in the scouring of embankments. A material point method–moving particle simulation coupling method was de veloped to predict the extent of such scouring. The effectiveness of the developed method was verified through reproduction analyses of the scouring of embankment caused by tsunami-induced overflow. Consequently, the developed coupling method could simulate the process and extent of scouring.

In the main shock of the 2016 Kumamoto earthquake, the rupture episode of the branching fault was inconsistent with conventional models. In the conventional models, rupture directions are assumed to be from major fault segments to their branches. In this study, we examined rupture patterns of branching fault systems by considering a variety of fault strengths and initial stress fields. In order to simulate the fault rupture including branching points, we proposed a new representation of enrich function for XFEM. The results suggest that ruptures can start from a branching point.

Classical solutions of bearing capacity are mainly based on the analytical solutions of metals. However,the tensile strength characteristics of soils are totally different from those of metals. In this article, no tensile stress condition for a first invariant of a stress tensor I₁ ≤0 is introduced and implemented into a rigidplastic finite element method. The influence of I₁ condition on the bearing capacity of shallow foundations are investigated by rigid plastic finite element analysis. The condition of I₁ ≤0 has a limited influence,except for the case of c(> 0), ϕ= 0. However, this condition contributes to the improvement of numerical accuracy and calculation stability of a rigid plastic finite element method.

Efficient treatments of correlations between material properties are important for the risk analysis of the ground because there are the correlations in soil. This paper proposes an approach to consider correlations between material properties in a stochastic elasto-plastic model with the spectral method. In the approach, the spectral expansion is performed on correlated parameters of material properties by applying the principal axis transformation. Increasing degrees of freedom is a practical problem when taking into account both spatial correlations and correlations between material properties. It can be avoided by considering spatial correlations of only the ?rst primary component of strongly correlated material properties. We simulate stochastic stress-strain relations of an elasto-plastic body in which the elastic modulus correlates to the strength. It is shown that the approach enables efficient evaluations of behaviors of elasto-plastic bodies with the correlated material properties despite the complex behaviors. It is difficult to predict the uncertainty of responses from one of the material properties without evaluations of the probabilistic yield process. Therefore, simulations considering correlations between material properties are essential for the risk quanti?cation of the ground.

The stabilized ISPH method is a particle method for incompressible flow problems, which is improved by a relaxation term in the pressure Poisson equation of ISPH method in order to maintain uniform particle distributions. It has been confirmed that the relaxation term contributes to enhancements of computational stability and volume conservation simultaneously. However, in fluid-solid multiphase simulations using the stabilized ISPH method, the corrections of the relaxation term become relatively large and negatively affect the computational results because of huge particle density errors around the fluid-solid surface. We propose a selective dual velocity ISPH method, which defines physical and transport velocities, individually. The physical and transport velocities are used for updating physical quantities and for particle positions, respectively, and only the transport velocity involves the affect of relaxation term. The selective dual velocity ISPH method is applied to a dam-break flow with solid particles to show the effectiveness of the proposed method by comparing with the stabilized ISPH method.

In this paper, the verification of a free vibration analysis of elasticity is discussed. An eigenvalue problem possessing solvable solutions is proposed for verification of the code and evaluation of numerical properties in real calculations. A homogenous and isotropic elastic material and cylindrical domain are employed and stationary modes of P-wave are considered as specific eigenmodes of vibration. A process to compare numerical solutions and exact ones proposed in this work is also presented and numerical properties of several finite element approximations are investigated.

Experimental testing revealed that the ground created the cone-shaped resisting area against the uplift of the belled pile. Then, the relationship between soil dilatancy characteristics and the uplifting capacity was investigated by using the numerical analysis which can simulate large deformations. The series of numerical simulation resulted in that the positive dilatancy under low confining pressure induced the truncated coneshaped resisting area in ground which increased the uplifting capacity.

In recent years, liquefaction damage of detached houses due to large-scale earthquake has become serious all over the world. If the existing small-scale structures such as houses are tilted due to liquefaction, they suffer more serious damage than they are subsided. In this paper, the effect of countermeasure against tilting damage due to liquefaction using floating grid-type improvement was investigated by 2D model tests under 1g condition using eccentric model structure, and non-linear FEM analyses. From the results, it was found that the floating grid-type improvement can significantly suppress the settlement damage of structure by suppressing the lateral flow of the liquefied ground even if the dead load of structure is eccentric. It was also found that lateral flow of the liquefied soil hardly occurs when the structure inclines due to liquefaction, and the effect of countermeasure against tilting damage due to liquefaction using floating grid-type improvement was not as high as the effect of suppressing subsidence damage. Therefore, it was found that desirable to use a floating grid-type improvement and drainage method to prevent tilting damage of structure.

This paper presents an analytical approach on the load carrying capacity of steel pipe beams with flange joints against boulder impact load by changing the impact position of the boulder. First, this proposed method was validated by comparing with the static and dynamic test results. Second, the load carrying capacities of steel pipe fix-ends beams with flange joints were investigated by comparing with the impact positions just above the flange joint (Type 1) with the one between flange joints (Type 2) using the elasticplastic FEM analyses under the static and impact condition. It was found that the case of ‘Type 1’ was more critical rather than ‘Type 2’ within the range of impact condition. This may be caused that ‘Type 1’ was not able to absorb the impact energy by the local deformation and dispated by the breaking energy of joint bolts. On the other hand, ‘Type 2’ was absorbed by the local deformation of a steel pipe and was not so critical than ‘Type 1’.

Scour occured by floods lead to large topography variation and serious damage. Mechanism of scouring have been evaluated energetically conventionally by bottom shear stress, but in late years it has been argued that fluid power let stress in ground fluctuate and promote scouring. In this research, We performed movable bed experiment and theoretical consideration compared with interpretation in past to quantify influence of seepage flow and pore water pressure change in ground under high speed fluid on scour. As a result, character of seepage behavior while scouring was greatly different from theory solution in porous fixed bed which scouring is not occured. Particularly, it clarified that seepage flow to drain away ground outside and liquefaction let stability of soil greatly fluctuate. In addition, arrangement about seepage behavior in outer layer which varies according to grain size and thickness of permeable layer can improve prediction precision of scour analysis and showed that it connects to efficient and effective maintenance.

This paper deals with a laboratory experiment of submarine landslides. In general, submarine landslides can break submarine cable and cause tsunami without any large seismic shaking. It has been thought that tsunami magnitudes by submarine landslides are strongly affected by the velocity of the sliding mass. Based on these backgrounds, our experiments were focusing on their trigger mechanisms due to excess pore water pressure. During the experiments, we measured the displacement and the velocity of the sliding mass. As a result, the experimental landslides were classified into three patterns based on time-profiles of the velocity. A velocity peak characterized patterns 1 and 2 at the initial period during the experiments, and those generated small slides. In contrast, pattern 3 developed into a large slide which behaved like a creep rupture; it slides slowly at the initial stage, reaches a steady-state with the constant velocity at the intermediate period, and slides rapidly at last. In addition, the relationship between the secondary creep velocity and the time to creep rupture can be linearly approximated on a double-logarithmic graph, as in the case of subaerial landslides.

This paper elucidates the reduction mechanism of an impulsive load of debris flow with driftwood utilizing the proposed distinct element method. In addition, the efforts were visualized contact force transmission mechanism inside the woody debris of analysis results by using con-tact force distribution illustration and velocity vector one. In the results, driftwood concentrated at the front part of woody debris is entrapped by an open sabo dam, and the proposed method was able to reproduce the load time relation obtained in the experiment. In addition, the driftwood functions as cushioning material, and the function reduced impulsive load of debris flow received from the subsequent flow of debris flow. It was found that the impulsive load of debris flow including driftwoods was smaller than that of bouldery debris flow.

Takayuki FUMOTO, Taijyu URA and Stephen A. HALL On the effective use of resources, the concrete should be made by mixing many kinds of material. Understanding the influences of the material qualities on properties of the concrete is necessary for a proper material mixture. One of the powerful tools for understanding those influences is a collaboration of an X ray computed tomography and a digital volume correlation for measuring displacement inside the concrete. We have been developing some techniques for applying to the concrete specimen. In this study, the strain distribution inside the specimens of concrete, paste and mortar with the coarse aggregate of different deformation resistance under the compressive stress were measured using our current techniques. As the results, it is cleared that our current technique shows the strain concentration area of a mortar part inside the specimen before a destruction. Moreover, It is necessary to further improvement of the current techniques for understanding the destruction process of mortar and concrete specimen in more detail.

In recent years, tornado and volcanic eruption have increased under the influence of climate change and crustal movement. In the events, collision of tornado missiles and volcano cinder with the important structures such as atomic energy facilities is concerned. This study conducted numerical analysis of the previous experiment conducted by the authors and investigated the mitigation effects by the aramid fiber sheets attached on the back surface of RC slabs. Numerical results reproduced the mitigation effect by the fiber sheet reinforcement and discussed the mitigation mechanism.

The strain space multiple mechanism model idealizes the behavior of granular materials based on a multitude of virtual simple shear mechanisms oriented in arbitrary directions. These mechanisms idealize the internal (or micromechanical) structure of granular materials with induced anisotropy and form a second-order fabric tensor, which relates macroscopic strain to macroscopic stress and has been updated to consider the effect of inherent anisotropy. This paper investigates the inherent anisotropy’s effect on the evolution of induced fabric caused by the change of macroscopic stress/strain. Comparison with Discrete Element Method (DEM) simulation demonstrates that the strain space multiple mechanism model accounting for inherent anisotropy has the potential to capture the essential features in the evolution of an induced fabric in granular materials.

This study aims to quantify the contribution of input parameters in granular flow simulations using DEM. A granular flow simulation with simple condition is performed using DEM under the different combination of the input parameters. Indexes that related to the traveling distance are defined, and surrogate models of the DEM simulation are built with the help of the radial basis function interpolation. The Monte Carlo Simulation is then carried out with quantified contribution rate of the input parameters. According to the results, the basal friction is the most important parameter, but the restitution coefficient strongly contributes characteristic of the front part. It was also found that the spring codfficient and friction angle between elements don’t have strong contribution.

The effects of the stress triaxiality on the dynamic mechanical characteristics of steel were not sufficiently investigated, even though the fracture strain of steel decreases with the increasing triaxial stress under static loading. In this study, dynamic tensile mechanical properties of SS400 steel were experimentally investigated by using a SS400 plate specimen with a notch to examine the effect of the stress triaxiality on the yield stress, tensile strength under high strain rate. Test results revealed that the dynamic strength properties of steel increased under dynamic loading, but the effect of stress triaxiality on the dynamic strength properties was not significant. The fracture strain of steel decreased with the increasing triaxial stress under high strain rate.

The phenomenon which the fatigue strength decreases even with the same weld joint type, due to an increase of the plate thickness or a combination of the plate thickness is known as a plate thickness effect. It is considered that stress concentration and stress gradient at the fatigue crack initiation position are factors of the plate thickness effect, and some experimental and analytical studies have been conducted. However, in the case of welded joints, it is not easy to clearly distinguish these influencing factors, and it is hard to say that the plate thickness effect is fully understood. The object of this study is to consider the thickness effect on the fatigue strength through FE analysis and comparison with the experimental results. As a result, the plate thickness effect observed in the experiment was able to be reproduced, and it was found that the crack initiation life was a major factor of the plate thickness effect observed in the experimental research.

Fatigue damage process is generally divided into two stages, fatigue crack initiation and propagation. Stress intensity factor or J-integral are usually used for driving force of fatigue crack propagation. The authors have recently proposed a fatigue crack propagation life assessment method based on elasto-plasticity response at a crack tip. Firstly, fatigue crack initiation life is assessed based on elasto-plasticity FEM and Δε-Nc criterion. Fatigue crack growth rate is calculated by da/dN = Δa/Nc, where Δa is a parameter that is introduced in the method and defines the fatigue crack growth rate. Although the previous study about 2D problem has shown the high applicability, the study has not focused on the surface crack problem or the effect of mean stress on fatigue crack propagation and threshold SIF. The purpose of the present study is to evaluate the fatigue crack propagation life against surface crack existing in the stress concentration field. Comparing with the experimental result, it was verified to evaluate accurately the fatigue crack propagation life in the depth and width direction of surface crack.

Similarities of hydraulic model experiments for river including seepage flow through saturated porous riverbed materials, are investigated here based on the governing equations of numerical model. With the theoretical and computational results, it is shown that more attention should be paid to keeping consistency with Froude number similarity. 1/N Scaled physical roughness heights of river bed are required to satisfy Froude similarity for river flow, 1/N scaled groundsill and grain size are required to satisfy Shields number similarity for bed load, 1/1 or 1/√N scaled permeability is required to satisfy partial similarity or Froude similarity as river flow. The similarities at boundary roughness and grain size will be lost due to a too small scale 1/N. and, a 1/1 scaled permeability under small scale 1/N experiment may lead the river flow reduces due to relative bigger seepage flow.

Experimental investigation on hydraulics in pipe flows connected to manhole yields that a stable flow in manhole is significant for the improvement of drainage faculty. The connection with a drop between inlet and outlet pipes at manhole and the installation of a control plate in manhole were proposed. In this case, the connection angle between inflow and outlet pipes at manhole was adjusted as 120 degrees. Also, the slope of the outlet pipe was settled as about 1/5 (Model II). For the inlet horizontal pipe flow, the discharge coefficient for Model II-2(including control plate) could be increased from 0.62 to 0.73 in comparison with that for Model II-1. Further, if the connection angle is changed from 120 to 180 degrees (Model III), the flow passing the inlet pipe entered mainly to the outlet pipe at the manhole, and the discharge coefficient could be increased from 0.62 to 0.777.

Analytical solutions of the governing equations of shoreline development are well-known and capable of solving many coastal problems regarding either natural morphology change or structure involvement. This study presents a new analytical solution of one-line model for a widely occurred phenomenon where the river mouth is gradually migrated to a new location after a dramatic change caused by overwhelmed disasters. This solution is in conjunction with the Heaviside function to overcome the challenge of time varying conditions. The application of this solution is proved by applying on an existent study site, Da Rang River mouth, Vietnam. The solution reveals the severe erosion triggered due to the graduall migration of the opening after a highly extensive flood. The decay process on the old river mouth position proves the erosion caused by the migration of the opening. The opening kept shifting to the left side of the river mouth until stabilizing after 6 years from the flood event.

Both level set methods and volume of fluid (VOF) methods are popular for computation of immiscible-fluid flows. Here a comparative study of the ACLS (accurate conservative level set) method, a level set variant, and the CICSAM (compressive interface capturing scheme for arbitrary meshes) method, a VOF variant, in terms of accuracy and efficiency has been carried out. Several benchmark tests revealed that the ACLS method combined with a high-order scheme shows a remarkable ability for interface capturing. While the CICSAM is available only at low-CFL-number conditions, its accuracy comparable to the ACLS method and easiness of implementation are also desirable. Application to the steady state air bubble test, which requires coupling with a Navier-Stokes equation solver, presented that a sophisticated method for calculation of interface normal and curvature is essential for VOF methods when surface tension force plays an important role.

Manholes connecting sewer pipes are important elements in an urban drainage system. The manholes exert impacts not only on the local flows in their vicinities but also on the sewer flows in the whole pipeline system. For an appropriate assessment of the capacity and efficiency of a drainage system, and to improve the accuracy of numerical models for urban flood simulation, it is crucial to understand the hydraulic properties of manholes. In this study, a series of laboratory experiments have been conducted with a scaled physical sewer system model to clarify the local energy losses and local flow structure in a junction manhole. In particular, the influences of the lateral flow discharge ratio, the lateral pipe alignment angle and the outlet pipe submergence condition on the local energy losses have been analyzed. In addition, the local three-dimensional velocity structure in the manhole has been characterized.

Gabion revetments have been installed in front of a beach scarp formed due to high waves as a measure against further scarp formation. Although gabion revetments are unable to completely protect the sand behind the structure unlike other inpermeable revetments, they induce natural recovery of the beach profile because of their permeability allowing cross-shore sand transport while protecting the eroded beach from wave forces. This study evaluate the performance of a gabion revetment formed by the round-shaped gabions so-called ’daruma-kago’ which have high frexibility in arranging a revetment and have higher long-term durability because of the efficiently packed rock fills. The round-gabion revetment shows sufficiently high stability against waves with an ordinary water level. The stability of the revetment in higher water level is found to be improved by changing the arrangement of the round gabions. Although porosity of the gabion and the gap between the round gabions allow sand to move through the gabion structure, the round gabions showed to slow down the development of a beach scarp. Also the round-gabion revetment was found to show more efficient beach-protection function if we enhance the sand protection function of the gabions by wrapping it with geotextile.

Alternate bars are regarded as wave phenomenon from its geometrical feature and propagation characteristics. The occurrence condition and development process of alternate bars was made clear in preceding study. However, its propagation which is important feature as wave phenomenon is often unclear. Especially, Alternate bars’s celerity is unclear whether they have spatial distribution and its temporal variation. In this study, in order to confirm the spatial distribution of alternate bars’s celerity and its temporal variation, we conducted firstly flume experiment, and measured the shapes of the water surface and the bottom. Secondly, we have qualified them with combination of measured values and estimation formula. As a result, we clarified that alternate bars ’s celerity has a spatial distribution, and that spatial distribution expands with time.

We have numerically examined turbulence transport processes in a convective boundary layer developed above a heated roughness, which mimics an urban canopy with buildings and streets by using large-eddy simulations (LESs). We considered two kinds of convective boundary layer flows: one was without an urban canopy and the other was with an urban canopy. The conditions of convective boundary layer and the configurations of urban canopy were set according to previous studies. An open source CFD code, OpenFOAM was used. The present simulation was based on a LES version with PIMPLEfoam of Open-FOAM, in which the option of RANS scheme was turn off. An one-equation subgrid-scale model was adopted to represent the subgrid-scale diffusion. The central differencing scheme and the 2nd-order back Euler scheme for time integration were chosen. Comparison with the LES for ideal case (without roughness) showed that the roughness enhanced the dissimilarity between thermal (temperature) and flow (velocity) fields, i.e., the vertical profiles of turbulence flux in the thermal fields significantly differed from those in the flow fields. The urban canopy yielded the increases in thermal turbulence fluxes, while the effects of urban canopy on the flow fields were quite small. Such differences were caused by differences of coherence structures developed above the urban canopy between thermal and flow fields, i.e., the thermal and flow fields were controlled by roll and streaky structures, respectively, which were visualized by contours of eiinstantaneous temperature and velocity fluctuations in horizontal planes.

The classification of sand waves and their indices have been studied separately for sand bars, dunes and ripples. The flow over the sand waves was empirically defined as shallow water flow on sandbars and non-shallow water flow on dunes and ripples. In addition, the results of experiments and measurements of surface waves on sand waves in rivers have already been published in previous studies. We have discussed the relationship between these classifications and indices, the aspect of flow pattern on sand waves, and the water surface waves caused by them from the fluid dynamic point of view. The present study is based on the small amplitude wave theory, and uses the depth-wavelength ratio to determine the aspect of flow over sand waves and we developed an unified index that allows us to examine whether or not a river is represented, and compared it with the results of real and experimental flumes. As a result, we were able to classify the sand waves and flow pattern in a unified manner using the depth-wavelength ratio.

In recent years, typhoons and heavy rains have caused enormous flood damages in various part of Japan. In this study, the evacuation simulations in which an agent grasps the flooding state based on the result of flooding analysis in advance from, for instance, a hazard map made by a local government are carried on. In consequence, it becomes possible that the agent completes evacuation in short time without been caught in the flooding by grasping the flooding state in advance. In this case, the rate of completion of evacuation is significantly higher than that when the flooding state are not known beforehand.

There is a plan to install MEMS accelerometers into many structures such as houses, road signs, etc. These sensors are able to observe a seismic response. Therefore, it might be a realistic idea to build a super high-density seismic network system by compiling the seismic data measured by the MEMS accelerometers into a database. However, it has a problem to treat a structural response as a seismic data because the structural response has relatively larger intensity than the ground motion. In this paper, the estimation method for the SI (Spectral Intensity) calculated from a structural response is examined. In this method, a transfer function between a structural response and a gournd motion is estimated based on a single degree-of-freedom system. The SI-value is estimated using the ground motion calculated from the structural response using the transfer function. A numerical simulation shows that the SI values estimated from a structural response has good agreement with the value estimated from the original ground motion.

Seismic fragility curve against slope failure with respect to peak ground acceleration (PGA) or instrumental seismic intensity (ISI) is constructed based on database of 4321 slople failures caused by Niigata-ken Chuetsu earthquake (2004). Since it is difficult to count the number of natural slope, the failure area ratio, which is the ratio of failure area to total area, is used for the fragility curve. Data of PGA and ISI at 131 sites are obtained from Japan Meteorological Agency and the National Research Institute for Earth Science and Disaster Prevention. Spatial distribution of seismic intensity is estimated by using Kriging based on the obtained data. The fragility curve is constructed by multiple regression analysis in which the objective variable is the failure area ratio, explanatory variables are slope angle and the seismic intensities. Failure area ratio can be obtained for arbitrary seismic intensity and slope angle by the proposed fragility cureve, e.g., the failure area ratio is around 0.1 when slope angle is 35 degree and PGA is 1350 gal or ISI is 6.3.